Evolution not Revolution: what can we learn about human health from nutritional ecology?

1

David Raubenheimer

Leonard P Ullman Chair of Nutritional Ecology

Faculty of Vet Science | School of Biological Science | Charles Perkins Centre

Nutritional Ecology is ….

• Ecological / evolutionary approach to nutrition

- focus on how nutrition mediates the relationship between animal & environment

- to determine health and wellbeing

Animal Biology

Environment

NUTRITION

health

lifespan

reproduction

etc.

health

lifespan

reproduction

etc.

health

lifespan

reproduction

etc.

Aims for the talk

NUTRITIONAL GEOMETRY

Animal Biology

Environment

NUTRITION

MODERN FOODS

HUMAN APPETITE

OBESITY

• Introduce Nutritional Geometry: approach for studying these interactions

• Show how it has been used to understand:

1) Obesity: a not so obvious question

5) Closer look at our environment

3) Stepping back in time: wild apes

6) A closer look at human biology

2) Nutritional geometry

4) Humans

1. Way back ….

• Staples

- lean meat

• Luxuries

- low GI vegetables

- honey - tubers

• Diet

- high protein (25–30%)

- fruit

(picture credits: Wikipedia)

• Grains

- high starch

• Domesticated fruit

- increased sugar

• Livestock

- higher fat than game • Diet

- reduced % protein

2. Agriculture

• Bulk extraction

- carbs - lipid

• Processed foods

• Diet

- further reduction in % P

3. Industrial revolution

Overview

- fats + carbs limiting - culture lifts limitation - satisfies appetite

- stone age physiology - modern environment MALADAPTED

• Problem

High protein High fats + carbs

• Solution

- evolved strong appetite for F+C

• New problem

The not-so-obvious question

WHY do we over-eat energy on diets high in fats + carbs?

1) Obesity: a not so obvious question

5) Closer look at our environment

3) Stepping back in time: wild apes

6) A closer look at human biology

2) Nutritional geometry

4) Humans

CA

RB

OH

YD

RA

TE

ga

in

PR

OT

EIN

ga

in

or C

AR

BO

HY

DR

AT

E

ga

in

PROTEIN gain

Nutrient space

- by constructing a multi-dimensional nutrient space

- models nutrition in terms of: (i) two or more NUTRIENTS

(ii) their interactive effects on animals/humans

What is nutritional geometry?

Can then plot in this space:

1. Nutrient requirements

2. Foods

3. Feeding

- and model the relationships among these

4. Consequences

etc..

1. Nutrient requirements

i. amount of nutrients

Ca

rbo

hyd

rate

Protein

Nutritional rail

0 0

•

2. Foods

low P:C

high P:C

intermediate P:C

ii. balance of nutrients

Protein

Carb

ohydra

te

Protein

i. balanced iii. complementary

- the animal gains nutrients in same balance as the food – as it eats, it “moves” along rail

ii. imbalanced

Protein

- what are the options?

•

3. Feeding

Carb

ohydra

te

Protein

deficit P

exce

ss C

Carb

ohydra

te

Protein

prioritise protein

prioritise carbohydrate

some balance of excesses & deficits

- there are main 3 options: - how to measure the rule of compromise?

“Rule of compromise”

e.g

. lif

esp

an

- response surface

4. Consequences

- or contour plot

Protein

Ca

rbo

hyd

rate

1) Obesity: a not so obvious question

5) Closer look at our environment

3) Stepping back in time: wild apes

6) A closer look at human biology

2) Nutritional geometry

4) Humans

How do they prioritise macronutrient intake?

?

Mountain gorillas Orangutans

• Record:

i. what is eaten

ii. how much

iii. nutrient content

• Analyse data using nutritional geometry

Method

• Direct observations

Jessica Rothman

1. Mountain gorillas

- mainly leaves: 8 months - mainly fruits: 4 months

Photos: J. Rothman

- low % protein - high % protein

Diet

• Rule of compromise

Biology Letters 2011 7:847-849.

Carb

s &

Fat

Kj/day

Available Protein Kj/day

leaf diet

35%

protein

silverbacks

adult females

juveniles

19% P 31% P

• Fruit diet: 19% P

i. regulated to target Carb + Fat intake

ii. to do so, over-ate P

fruit diet

• Leaf diet: 31% P

Results

- i.e. prioritise non-P energy

Closer look at what it means Low P H

igh P

• Similar carb + fat intake on all diets

• AND higher P intake on high-P diets

• i.e. Mountain gorillas eat MORE energy on high P diets

• Unlikely to get fat in modern human environments

- 49 orangs

- Observations over a 7 year period

- 2,233 full day observations

- 49000 hours

Erin Vogel

2. Orangutans

- also physiology

Method

• Preference

Diet

- but availability unpredictable

• So also eat

- leaves + cambium

- fruits: high carbs + fat high carbs + fat

hig

h p

rote

in

- high protein

% Trees fruiting

Protein intake (Kcal) Carb

+ fat

inta

ke (

Kca

l)

Result

• Wide variation in carb + fat intake

• Tight regulation of protein

- i.e. prioritise non-P energy

Implications Low P

Hig

h P

• Will over-eat fats + carbs on low-P diets

• Or under-eat fats + carbs on high-P diets

• Could the same pattern explain why humans get fat in modern environments?

% Trees fruiting

Protein intake (Kcal)

Carb

+ fat

inta

ke (

Kca

l)

And …..

laying down fat

burn fat

• Physiology tells us

- lay down fat when energy (fruit) is abundant

- draw on it when energy is scarce

• i.e. Adapted to “boom and bust” ecology

• As are humans!

1) Obesity: a not so obvious question

5) Closer look at our environment

3) Stepping back in time: wild apes

6) A closer look at human biology

2) Nutritional geometry

4) Humans

Two Questions:

i. How do we compare with other primates?

ii. What are the implications for obesity?

- Prioritise P, like orangs

- Meta analysis: 26 published trials

model

Gosby et al., Obesity Reviews 2014

Alison Gosby Human macronutrient regulation

- 3 experimental studies: Oxford, Sydney, Jamaica

Protein

Carb

ohydra

te +

fat

14% P 12.5% P

- A small change in % P in foods will result in a disproportionately large change in the amount Carbs + Fat eaten

- For example, a 1.5% decrease in % P

14% increase in C + F eaten

Implications of protein prioritisation for obesity

- could this explain the obesity epidemic?

And weight loss …

-> Called the “Protein Leverage Hypothesis” (PLH)

- a small increase in % protein will result in a large decrease in carbohydrate and fat eaten

- could this explain why high protein weight loss diets work?

Protein

Ca

rbo

hyd

rate

+ fa

t

14% P 15.5% P

(1.5%)

11% decrease in C+F eaten

0 10 20 30 40 50 600

20

40

60

80

Energy intake (kj/day)

Protein (% kj)

Fa

t (%

kj)

5000

6000

7000

8000

9000

10000

11

00

0 [0]

[20]

[40] [60]

(10:30:60)

1) Energy intake increases with decreasing %P

Two predictions of PLH

- Yes!

increasing energy

Prediction 2). Dietary % P has decreased with the rise in obesity

Data: FAOSTAT 2010

- e.g. USA

- yes!

1) Obesity: a not so obvious question

5) Closer look at our environment

3) Stepping back in time: wild apes

6) A closer look at human biology

2) Nutritional geometry

4) Humans

The general question

• PLH suggests that because humans prioritise P, energy intake will be higher on low-P diets

• What changes in the environment cause the shift to low-P diets?

??

A Role for Economic$

- obesity is more common among low SES groups

• Is this related to variation in dietary % protein?

Source: Australian Institute of Health and Welfare

• A clue that economics is involved:

0 10 20 30 400

20

40

60

80

100

$US

Protein g/100g

Fa

t +

ca

rbo

hyd

rate

g/1

00

g

0.6

0.8

1

1.2

1.4

1

.6

1.8

2

2.2

1) Protein is more expensive than fats and carbs

$/100g

Brooks et al. (2009) Obesity reviews.

Three predictions

- 106 supermarket foods

- price increases with P

- not with fats & carbs

• Test

• Result

- compared the separate contributions of each g of Pro, Fat and Car to the cost food

- suggests an economic incentive to eat low P foods

• Conclude

cheap expensive

- = high energy intake

Prediction 2). Low SES groups eat low-protein diets

- 14 diet surveys of low SES indigenous Australian communities

- most have low % P relative to recommended range

• Aboriginal study

• Low relative to:

[With Aboriginal Nutrition Project Node]

1. Australian recommendations (AMDR)

students

pregnant women

2. Higher SES groups

Prediction 3). Low protein diets are associated with high energy intake

pregnant women - yes!

- as we found in experimental studies

students

A role for global change

Raubenheimer et al. (2013), in review

[With Human Food Chain Project Node]

Rapid increase in atmospheric CO2

- CO2 enrichment results in an average 54% increase in the carbohydrate:protein ratio

Raubenheimer et al. (2014), in review

0 10 20 30 40 50 600

20

40

60

80

Energy intake (kj/day)

Protein (% kj)

Fa

t (%

kj)

5000

6000

7000

8000

9000

10000

11

00

0

• Robinson et al. 2012*

Alters plant composition

*New Phytologist 194:321-336

Likely to impact not only on plant foods

Ta

rge

t ra

tio

+ 54%

+ 22%

1) Obesity: a not so obvious question

5) Closer look at our environment

3) Stepping back in time: wild apes

6) A closer look at human biology

2) Nutritional geometry

4) Humans

Protein

Ca

rbo

hyd

rate

+ fa

t

A Role for Protein Requirements

- protein leverage is expected to be proportional to protein requirements

- can variation in P requirements help explain variation in obesity?

1. Obesity as a cause of increased P requirements

• Physiological effect of obesity: Protein efficiency is reduced

• So more protein needs to be eaten to meet needs

• Suggests a positive feedback: obesity drives further excess energy intake and obesity

http

://ww

w.th

eguard

ian

- and P leverage is enhanced

2. Dietary history (evolutionary or developmental?)

• For example, Alaskan Inuit

• Populations with high protein traditional diets are particularly susceptible to obesity when transferring to westernised diets

- among the highest rates of obesity globally

- traditional diet = 30-35% P

• Could this explain their susceptibility to obesity?

Data from Kang-Jey et al. (1972), AJCN 25:737-745.

- have developed low P efficiency

- so higher P intake need

3. Effects of early nutrition (“developmental programming”)

• Formula-fed infants are more susceptible to obesity later in life than if breast-fed

• Milk formula is much higher in protein than human breast milk

breast

infant

follow-up

• Could high-P early in development lead to reduced P efficiency hence increased P target and energy intake?

Data from Koletzko et al. (2009), AJCN 89:1836-49.

CONCLUSIONS

• Nutritional geometry can help to measure interactions among appetites for different nutrients

• Helps to understand the mismatch between humans and modern food environments

• And identify new areas for research and intervention

- factors that reduce % protein in the human diet

- factors that decrease human protein efficiency

Thank you for your attention!